Harlequin Duck Histrionicus Histrionicus Distribution and Stonefly

Harlequin Duck Histrionicus histrionicus distribution and stoneﬂy nymph availability in the Maligne Valley watershed – a preliminary study

N.A. McCutchen1 & R.C. Ydenberg2

1Department of Biological Sciences, University of Alberta, Edmonton, AB, T6G 2E9, Canada. Email: [email protected]

2Behavioral Ecology Research Group, Department of Biological Sciences, Simon Fraser University, Burnaby, BC, V5A 1S6, Canada. Email: [email protected]

The relationship between Harlequin Duck Histrionicus histrionicus distribution and stoneﬂy (Plecoptera) nymph availability in the Maligne Valley watershed, Jasper National Park, Canada, was investigated from 1999 to 2000. Invertebrate sampling and duck counts were concentrated in the Maligne Lake Outlet (MLO), Middle Maligne River and Lower Maligne River. Harlequin Ducks were more likely to use the two sites with high nymph availability (MLO and Lower Maligne) than the site with low nymph availability (Middle Maligne). The relationship between Harlequin Duck distribution and other invertebrates was less consistent. These results suggest that stoneﬂy nymphs play an important role in the distribution of Harlequin Ducks in the Maligne watershed.

Key Words: Histrionicus histrionicus, Plecoptera, pre-breeding foraging

Introduction nymphs seem particularly important to the distribution of Harlequin Harlequins Ducks Histrionicus Ducks (Hunt 1998; McCutchen 2001), histrionicus are Holarctic sea ducks although detailed diet studies have that migrate inland in spring to breed not been conducted. These nymphs on rivers and streams after wintering are relatively large, common and on coastal waters. Breeding sites conspicuous invertebrates that are for the Pacific population are found available throughout the Harlequin throughout the Western Cordillera and aDuck breeding season (McCutchen include the Maligne Valley watershed 2001). in Jasper National Park, Canada (Hunt In a previous study, Hunt (1998) & Ydenberg 2000 and references suggested that Harlequin Duck feeding therein). Harlequin Ducks feed in sites depended on the availability of the lower reaches of the watersheds stonefly nymphs. The main objective before egg laying and incubation (Hunt of this study was to further investigate & Ydenberg 2000). Use of feeding sites this link, although small sample sizes varies spatially and temporally. For and limited spatial and temporal scope example, Harlequin Duck use of the still render the study correlative. Maligne Lake Outlet (MLO) is typically Nonetheless, if stonefly nymphs are high whereas use of the Middle Maligne key to the distribution of Harlequin River is rare (Hunt & Ydenberg 2000). Ducks, Harlequin Ducks should use In years when use of the MLO is low, sites with high nymph availability (as Harlequin Ducks increase their use of determined by biomass/m2) more than other feeding sites, including the Lower sites with low nymph availability. This Maligne River (McCutchen 2001). study is important for Harlequin Duck Prey availability may account for management in Jasper National Park, at least part of the variation in the where recent variance in Harlequin distribution of Harlequin Ducks along Duck numbers at the MLO has raised the Maligne (Hunt 1998). The importance concerns about the stability of the local of prey availability to Harlequin duck population (Hunt 1998). However, Duck productivity has been shown this study will not only provide insight previously (Bengtson & Ulfstrand 1971; into Harlequin Duck ecology in the Gardarsson & Einarsson 1994; Rodway Maligne watershed. Harlequin Ducks 1998a; Wright et al. 2000) and ultimately are a poorly studied species, despite suggests that Harlequin Ducks that concern about their status in North feed in areas with the most prey have America (COSEWIC 2001). Therefore the highest reproductive success. any information on the ecology of this The Harlequin Duck diet is composed species will be useful for supporting mainly of aquatic invertebrates that relevant conservation actions. are found on and under rocks on the river bottom (Bengtson & Ulfstrand Methods 1971; Gardarsson & Einarsson 1994; Rodway 1998a, b; Wright et al. 2000). Study Area – This study was conducted In the Maligne, stonefly (Plecoptera) in the Maligne Valley watershed Harlequin Duck in the Maligne Valley 77 in Jasper National Park, Alberta, and then into the Middle Maligne River Canada (Figure 1). This 70-km-long, via the MLO. Most of the glacial silt glacially fed watershed includes two settles in this large lake before it enters lakes (Maligne and Medicine Lake) the Middle Maligne. The Middle Maligne and three main river sections (Upper joins Medicine Lake approximately 15 Maligne, Middle Maligne and Lower km downstream of the MLO. A karst Maligne). The valley bottom contains system of caves connects Medicine Lake mixed-species forests dominated by to the Lower Maligne, which eventually Lodgepole Pine Pinus contorta, White drains into the Athabasca River. This Spruce Picea glauca, Engelmann Spruce water system is part of the Mackenzie Picea engelmannii and Trembling Aspen watershed, which eventually empties Populus tremuloides. Average annual into the Arctic Ocean. Rainbow Trout water discharge (i.e. rate of flow) at the Oncorhynchus mykiss and Brook Char MLO is 15.9 m3/s. Salvelinus fontinalis were introduced Glacial meltwater drains from the into Maligne Lake in the 1920s and 1960s Upper Maligne River into Maligne Lake and are now universally distributed

Figure 1. The Maligne Valley watershed, Jasper National Park, Alberta, Canada. 78 Harlequin Duck in the Maligne Valley between Maligne and Medicine Lakes for ducks. Surveys were made every (Sullivan 1989). No fish populations week from the beginning of May to the are established in the Lower Maligne end of June. as the karst system of caves upstream Surveys started at 0600 h at the and the Maligne Canyon downstream bottom of the watershed (i.e. the act as barriers to colonisation (Sullivan Lower Maligne) and usually ended 1989). at 1200 h. At each site, the upstream Harlequin Duck counts – This study and downstream sections of the concentrated on Harlequin Duck river/outlet were searched for birds. use of the MLO, Middle Maligne and All birds were counted and identified Lower Maligne (Figure 2). The MLO is to species. Banded Harlequin Ducks traditionally used as a ‘club site’ (i.e. were identified where possible, and a relatively large gathering of non- this made it possible to determine territorial birds) by up to 12 Harlequin whether the ducks were moving up Ducks while the Lower Maligne is and down the watershed during the defended by a few territorial pairs survey. Weekly counts were averaged (Hunt & Ydenberg 2000). Harlequin to yield an estimate of mean Harlequin Ducks are rarely observed on the Duck numbers in each section of the Middle Maligne River (Hunt & Ydenberg watershed. Visual observations were 2000). In 1999 and 2000, Harlequin aided by a telescope and binoculars. Duck distribution in the watershed Prey availability – Invertebrate was measured using surveys made biomass (mg/m2) on the substrate at 38 sites along the Maligne Valley surface was used to estimate prey Road. This road runs parallel to the availability in the MLO, Middle and Lower and Middle Maligne and ends Lower Maligne in 1999 and 2000. at a bridge that crosses over the MLO. Surber and Hess sampling, which are Consequently, most of the two rivers more traditional methods of sampling and all of the MLO could be surveyed substrate invertebrates, could not be

Figure 2. Map of the sampling sites in this study. Site 1 refers to the Maligne Lake Outlet, Jasper National Park, Alberta, Canada. Site 2 was located about 5 km downstream of the MLO, while Site 3 was located about 8 km downstream of Medicine Lake. Harlequin Duck in the Maligne Valley 79 used in this study because the surface (see McCutchen 2002 for detailed substrate was too large (i.e. medium to methods for obtaining this function). large rocks interspersed with boulders, Biomass estimates for each sample Hunt 1998; McCutchen 2002). Instead, were log transformed (i.e. ln(x+1)) to invertebrates were collected using normalise the distributions (e.g. Elliott ‘five-rock’ sampling: an aquatic D-net 1977). Invertebrate biomasses were was positioned downstream of a rock compared among sites in 1999 and 2000 (medium sized, 8–12 cm wide). The using two-way ANOVAS in which site and rock was lifted into the net and all the year were the main factors (α = 0.05). invertebrates on and around the rock All analyses are based on transformed were washed into the net. The volume data, although back-transformed of the rock was then estimated using means are also provided. water displacement. This procedure was repeated four more times. Rocks Results in the sample were approximately 1m apart and selected randomly as Harlequin Ducks arrived in the the collector moved progressively Maligne watershed by mid-May in upstream. The total sample was 1999 and 2000, generally increasing in transferred to a sampling vial and number until June (Hunt & Ydenberg preserved in a 70% ethanol solution. 2000; McCutchen 2002). Based on Sampled rocks were not returned to the ring readings, individual Harlequin river. This procedure was repeated five Ducks did not move up and down the times to give five ‘five-rock’ samples watershed during any of the surveys. per site each sampling day. Replicates Most of the ducks were observed on were about 10 m apart. Samples were the MLO (Table 1). Use of the Middle collected once or twice a week from Maligne was rare whereas use of the the beginning of April to the end of Lower Maligne was low but consistent. June. Collection areas within each of Stonefly nymph biomass was higher the three sites were restricted to areas in the Lower Maligne and MLO than in where Harlequin Ducks had previously the Middle Maligne in both years (site: been observed. F2, 394 = 24.8, P < 0.001; year: F1, 394 = In the laboratory, invertebrates 0.2, n.s.; site*year: F2, 394 = 0.7, n.s.). were sorted into two groups using The biomass of other invertebrates Clifford (1991): stonefly nymphs and decreased with distance upstream other invertebrates (mainly mayfly towards the MLO (site: F1, 394 = 84.4, (Ephemeroptera) nymphs, blackfly P < 0.001; site*year: F2, 394 = 3.3, (Diptera) larvae, and caddisfly n.s). This trend was consistent across (Trichoptera) larvae). Samples were both years (F1, 394 = 2.2, n.s.). In all dried for 24 h at 80 °C and weighed sites, other invertebrates were more with an Ohaus Precision Standard abundant than stonefly nymphs. Total Balance (accuracy to 1 mg). Total rock invertebrate abundance was highest volume for each sample was converted in the Lower Maligne and lowest in the to surface area, where surface area Middle Maligne. = 13.875 . log volume3.603, r2 = 0.97 80 Harlequin Duck in the Maligne Valley

Table 1. Harlequin Duck abundance, mean stoneﬂy nymph biomass and mean other invertebrate biomass in the Lower Maligne and Middle Maligne River and Maligne Lake Outlet, Jasper National Park, Canada, in 1999 and 2000. Harlequin Duck abundance is represented as percentage use (counts are in parentheses). Invertebrate biomasses ± 95% conﬁdence intervals are log transformed (untransformed data are in parentheses).

Measure Lower Middle MLO

1999 % Harlequin Ducks 13 0 87 (2) (0) (13) Stoneﬂy nymph biomass (mg/m2) 1.3±0.2 0.7±0.3 1.4±0.3 (5.9±2.4) (6.1±4.1) (9.1±4.3) Other invertebrate biomass (mg/m2) 3.7±0.2 2.9±0.2 2.5±0.2 (50.3±7.2) (24.2±4.5) (17.9±3.5)

2000 % Harlequin Ducks 8 8 84 (4) (4) (42) Stoneﬂy nymph biomass (mg/m2) 1.4±0.3 0.5±0.2 1.6±0.3 (13.6±14.0) (2.8±2.7) (9.2±3.4) Other invertebrate biomass (mg/m2) 3.7±0.2 3.0±0.2 2.1±0.3 (49.3±9.7) (26.8±7.3) (13.3±4.2)

Both years combined % Harlequin Ducks 11 4 85 Stoneﬂy nymph biomass (mg/m2) 1.4±0.2 0.6±0.2 1.5±0.2 (9.3±6.3) (4.6±2.5) (9.1±2.8) Other invertebrate biomass (mg/m2) 3.7±0.1 2.9±0.2 2.3±0.2 (49.9±5.8) (25.4±4.1) (15.9±2.8)

Discussion stonefly nymph prey. The relationship between Harlequin Ducks and other This study suggests that the distribution invertebrates was less consistent. For of Harlequin Ducks in the Maligne example, Harlequin Ducks were less watershed is most closely associated likely to use the Lower Maligne than with stonefly nymphs. Harlequin the MLO despite its having the most Ducks were predominantly found in total invertebrates, the most other the MLO and the Lower Maligne, which invertebrates and the second highest are sites with relatively high stonefly biomass of stonefly nymphs. The nymph biomass. Harlequin Ducks were biomass of other invertebrates was rarely observed in the Middle Maligne, also higher in the Middle Maligne than which is a relatively poor source of in the MLO. Harlequin Duck in the Maligne Valley 81

These findings are consistent affect Harlequin Duck habitat use with the results of previous work in (e.g. pH, temperature, river width, the Maligne watershed (Hunt 1998; predators, etc.; e.g. Rodway 1998b). Hunt & Ydenberg 2000). It is not Such research is necessary before known why nymphs are so scarce in more definitive conclusions about the Middle Maligne, although Brook Harlequin Duck–prey interactions can Char and Rainbow Trout use it as a be made. However, these results are nursery (Sullivan 1989). Predation suggestive of a relatively strong link by these introduced fish may keep between Harlequin Ducks and stonefly nymph biomass low, and, as a direct nymphs, warranting further research consequence, render the Middle in the Maligne and in other Harlequin Maligne a poor feeding site for Harlequin Duck breeding sites where Harlequin Ducks. Low, though consistent, use Duck–prey relationships remain of the Lower Maligne is harder to unknown. Such information may be explain. However, nymphs in the Lower critical to the conservation of local Maligne are primarily herbivorous (e.g. Harlequin Duck populations. Nemouridae) while those in the MLO are primarily predatory (e.g. Perlidae) Acknowledgements (McCutchen 2001). In other studies, the presence of predatory stoneflies The authors wish to acknowledge results in higher rates of fish growth Ross Vennesland, Jessica Macdonald, and reproduction (Soluk 1993; Soluk Chantel Hamil, Jamie McMillan and & Richardson 1997). It is possible that D., L. and Bub McCutchen for help a similar relationship is occurring in in the field. We also wish to thank the Maligne watershed, ultimately Jasper National Park, Superintendent meaning that the MLO is the best place Ron Hooper and Park Wardens Mike to forage. Nymphs also influence the Wesbrook, Mike Comeau, Ward Hughes distribution of breeding Harlequin and Pete Clarkson for supporting this Ducks in Labrador (Rodway 1998b), project. Support for field work was though work in other countries suggest provided by the Challenge Grants in that blackfly and caddisfly larvae are Biodiversity Program (supported by the more important (Bengtson & Ulfstrand Alberta Conservation Association) and 1971; Gardarsson & Einarsson 1994; the Natural Sciences and Engineering Wright et al. 2000). These differences Research Council of Canada. Personal suggest that Harlequin Duck feeding funding for first author was provided by ecology cannot be generalised because the Natural Sciences and Engineering it is system-specific. Research Council of Canada, Simon The evidence from the present study Fraser University, and the Canadian is correlative, limited to small sample Fishing Company. sizes and based on one watershed in the Rocky Mountains. Consequently, the results should be interpreted with caution. Moreover, the study did not investigate other factors that could 82 Harlequin Duck in the Maligne Valley

References McCutchen, N.A. 2002. The response of stonefly (Plecoptera) nymphs to Bengtson, S-A. & Ulfstrand, S. 1971. seasonal increases in predation risk. Food resources and breeding Canadian Journal of Zoology 80: 967- frequency of the Harlequin Duck 972. Histrionicus histrionicus in Iceland. Rodway, M.S. 1998a. Activity patterns, Oikos 22: 235-239. diet, and feeding efficiency of Clifford, H.F. 1991. Aquatic Invertebrates Harlequin Ducks breeding in northern of Alberta. University of Alberta Press, Labrador. Canadian Journal of Zoology Edmonton, Alberta, Canada. 76: 902-909. COSEWIC 2001. Species at Risk in Rodway, M.S. 1998b. Habitat use by Canada. Committee on the Status of Harlequin Ducks breeding in Hebron Endangered Wildlife in Canada. URL: Fiord, Labrador. Canadian Journal of http://www.cosewic.gc.ca/eng/ Zoology 76: 897-901. sct5/index_e.htm (Last Updated: Soluk, D.A. 1993. Multiple predator May 2001). effects: predicting combined Elliott, J.M. 1977. Some Methods for functional response of stream fish the Statistical Analysis of Samples and invertebrate predators. Ecology of Benthic Invertebrates, 2nd edn., 74: 219-225. Scientific Publication No. 25. Soluk, D.A. & Richardson, J.S. 1997. The Freshwater Biological Association, role of stoneflies in enhancing growth Ambleside, England; 160 pp. of trout: a test of the importance of Gardarsson, A. & Einarsson, A. predator–predator facilitation within 1994. Responses of breeding duck a stream community. Oikos 80: 214- populations to changes in food supply. 219. Hydrobiologia 279/280: 15-27. Sullivan, M.G. 1989. Sport Fishery Hunt, W.A. 1998. The Ecology of Harlequin on the Maligne River during August Ducks Histrionicus histrionicus 1989. Fish and Wildlife Division, Breeding in Jasper National Park, Alberta Forestry, Lands and Wildlife, Canada. Unpublished MSc thesis, Edmonton, Alberta, Canada; 30 pp. Simon Fraser University, Burnaby, Wright, K.K., Bruner, H., Li, J.L., Jarvis, BC, Canada. R. & Dowlan, S. 2000. The distribution, Hunt, W.A. & Ydenberg, R.C. 2000. phenology, and prey of Harlequin Harlequins Histrionicus histrionicus Ducks, Histrionicus histrionicus, in a in a Rocky Mountain watershed. 1: Cascade Mountain stream, Oregon. Background and general breeding Canadian Field Naturalist 114: 187- ecology. Wildfowl 51: 155-168. 195. McCutchen, N.A. 2001. Temporal and Spatial Patterns of Stonefly (Plecoptera) Nymph Abundance Relative to Predation Risk in the Maligne River, Jasper National Park, Canada. Unpublished MSc thesis, Simon Fraser University, Burnaby, BC, Canada.